A FRAMEWORK FOR DEFINING SIMULATORS WITH WHICH

TO TRAIN GLOBAL SOFTWARE DEVELOPMENT

Miguel J. Monsor, Aurora Vizcaíno and Mario Piattini

Alarcos Research Group, Institute of Information Technologies & Systems, Escuela Superior de Informática

University of Castilla-La Mancha, Paseo de la Universidad 4, 13071, Ciudad Real, Spain

Keywords: Global Software Development, Distributed Software Development, Engineering Education, Educational

Environment, Teaching Model, Simulators, Virtual Agents.

Abstract: The teaching and training of Global Software Development (GSD) entails several well-known difficulties,

of which the problem of establishing environments in which students can learn by practicing in realistic

scenarios is commonly reported. In this paper we propose an educational framework that uses simulation to

train future participants to confront the principal problems encountered in GSD (cultural, language and

communication problems). Our framework therefore provides support for the design of these simulations by

means of a tool that trains its users in the typical problems that may occur during interactions between

distributed members. The simulations place learners in realistic GSD scenarios in which they will interact

with virtual participants, thus permitting them to confront the collaborative, organizational and technical

problems of GSD.

1 INTRODUCTION

Global Software Development (GSD) is currently

being addressed in educational environments with

the aim of training software engineering students in

the challenges that it entails. The principal problems

of GSD are specifically related to the establishment

of an effective communication between the

distributed participants (Monasor et al., 2009) who

must interact with people from different cultures and

different time zones in a common language in order

to jointly develop a software project.

Participants in GSD activities therefore require

additional skills in order to minimize the impact of

the inexperience and fears that they frequently have

to confront when dealing with problematic GSD

situations (Casey and Richardson, 2008), and to

avoid costly delays in time-to-market.

Behaviour in GSD as regards communication is

different to that encountered in other environments

(Cemile et al., 2009). The ability to persuade another

and the willingness to cooperate decrease with

distance, leading to a common deception of the team

members (Bradner and Mark, 2002). It is therefore

essential for students to develop the following set of

skills (Monasor et al., 2010):

• Ability to communicate effectively using a

common terminology and language.

• Performance in the use of synchronous and

asynchronous means of communication.

• Informal communication and improvisation

skills.

• Knowledge of language, cultural and ethical

issues.

• Leadership and conflict resolution skills.

• Time management skills.

• Managing ambiguity and uncertainty. Ability to

evaluate information critically.

• Skills to gain the interlocutor’s confidence and

trust.

• Knowledge of negotiation skills and contract

writing in a common language.

• Collaborative work skills.

However, it is difficult for educators to prepare

students in these sorts of skills since they have to

manage distributed activities in collaboration with

distant institutions. Moreover, it is not easy to

reproduce the conditions of real GSD developments,

principally because of the resources required and the

time limitations of the courses.

We present a framework whose objective is to

provide theoretical and practical lessons that will

261

J. Monsor M., Vizcaíno A. and Piattini M..

A FRAMEWORK FOR DEFINING SIMULATORS WITH WHICH TO TRAIN GLOBAL SOFTWARE DEVELOPMENT.

DOI: 10.5220/0003459902610264

In Proceedings of the 6th International Conference on Software and Database Technologies (ICSOFT-2011), pages 261-264

ISBN: 978-989-8425-77-5

 2011 SCITEPRESS (Science and Technology Publications, Lda.)

allow students to acquire communicative and

teamwork abilities through the simulation of

multicultural GSD environments.

In order to provide support for these activities,

we have developed VENTURE (Virtual

ENvironment for Training cUlture and language

problems in global softwaRe dEvelopment), a virtual

training environment that places students in a

simulated GSD scenario in which they interact with

Virtual Agents (VAs) from different cultures that

simulate realistic experiences. These simulations are

carried out by using written communication tools,

such as e-mail and instant messaging.

2 VENTURE

VENTURE consists of a platform integrated into an

e-learning system that supports the GSD educational

framework presented.

The innovative value of this framework lies in its

rigorous support for training, in that it not only

copes with cultural and linguistic differences, but

also improves attitudes for collaborative group work

without the need for real partners. The aim is to

provide theoretical lessons and simulated practices

in GSD, supported by a tool that simulates realistic

GSD collaborative environments. Teachers provide

students with the theoretical lessons on GSD

activities in class, and students can then go on to use

VENTURE to execute the training scenarios in the

laboratory.

VENTURE simulates GSD virtual meetings in

which students interact with VAs using a common

language (usually English). VAs communicate with

students textually and in an autonomous manner in

order to allow them to confront cultural and

linguistic problems similar to those that appear in

real environments.

Learners are placed in a virtual scenario and

work on the resolution of certain GSD activities by

interacting with VAs of different cultures. Students

can play the different roles in the process of GSD by

interacting with the VAs. The interactions are

guided by a Virtual Colleague (VC), which is

another kind of VA that will also correct the

students’ inappropriate interventions.

VENTURE introduces students to the

characteristics of the project to be developed and

their role. They are also given a software

engineering task or a responsibility that they must

accomplish.

Each lesson has practical materials associated

with it that must be delivered at prescribed

milestones. The students will generally have to

complete a document, or develop software in

accordance with the purpose of the scenario. They

must also answer a questionnaire at the end of the

course.

The virtual meetings are designed to reflect the

typical problematic or controversial situations

encountered in GSD, and the students are therefore

encouraged to find a solution to the problems by

interacting in the correct manner.

The definition of these meetings is based on

VTRML (VenTuRe Markup Language), which is an

extension to the XAML (Extensible Application

Markup Language) and permits the definition of all

the elements required, including the cultural and

linguistic rules for the scenario. The principal

objective of this is to improve the students’ skills

and performance when dealing with cultural

problems during the simulated conversations.

3 ARCHITECTURE

OF VENTURE

VENTURE permits the design and development of

GSD simulators by means of a client-server

architecture which is presented in Figure 1.

On the server side, an e-learning application

provides students with the course resources and the

facilities needed to carry out the training activities.

This part is made up of the following modules:

Resource Repository (1): This contains the

theoretical lessons assigned. Each theoretical lesson

is formed of virtual meetings which can be executed

at any time.

Task Area (2): A repository including the practical

activities assigned, in which the delivery deadlines

are specified. In this area, it is also possible for the

students to upload deliverables and review the

evaluation and instructor’s comments for these

activities.

Forum and Wiki Module (3, 4): This is used by the

instructors to include global communications, and by

the students to interact with instructors and partners.

Evaluation Area (5): Continual exams and

questionnaires which serve to evaluate the students’

learning.

On the server side, the theoretical material is

stored in the Pedagogical Module (6) and is

structured with reference to the knowledge areas:

software requirements, software design, software

construction, software testing, software quality,

ICSOFT 2011 - 6th International Conference on Software and Data Technologies

262

Figure 1: Venture architecture.

software maintenance, configuration management,

software engineering management and software

engineering process.

The cultural problems database (7) consists of a

repository that contains the set of cultural problems

and recognized differences that might affect

communication in GSD scenarios. The linguistic

problems that can appear when participants interact

textually with a non-native language are stored in

the language problems database (8). The linguistic

rules considered are classified according to the kind

of problem that they deal with and include any

relevant information that may be useful for

correcting the students’ actions. The information

contained in both the cultural database and the

linguistic database is managed by the Rules Editor

interface (9), which is made available to the

instructors through its cultural management module

(10) and language management module (11).

The skills required in GSD are stored in the

database (12) which contains best practices for

training the skills needed in GSD.

The VA profile (13) can be managed through the

VA profile management module (14) which permits

new characters to be included or existing ones to be

modified, and also allows them to be incorporated

into the theoretical materials or practical scenarios.

The Workflow Engine (15) is responsible for

executing the meeting workflows by reading the

definition of the meeting, and orchestrating the

sequential execution of the corresponding phases.

The engine additionally makes it possible to save

the log through the login unit (16) of the

conversation, so that the instructor can review it

later.

This workflow engine reads all the information

related to the simulation and makes the appropriate

transformations for each phase in order to generate

the AIML language. It does this by using the

transformation unit (17) to obtain information that is

understandable by the chatbot system (18), in the

case of synchronous interactions, and the Email

analyzer (19), in the case of asynchronous

interactions. The Evaluation unit (20) serves as a

tracking mechanism of the students actions to inform

both students and instructors of their skills and

results from the use of this framework, providing a

continuous and real-time evaluation.

The Workflow Designer (21) is a graphical tool

that the course designers use to define and modify

the virtual meetings. The virtual meetings are

structured as sequential workflows made up of a set

of phases containing the specifics details of the

conversation for that phase.

4 CONCLUSIONS AND

RESEARCH AGENDA

The chief advantage of using VENTURE is that

students are more independent as they do not need to

interact with distant partners and they can train at

any moment without depending on the availability of

other partners or colleagues. Students can also play

different roles while interacting with VAs and learn

A FRAMEWORK FOR DEFINING SIMULATORS WITH WHICH TO TRAIN GLOBAL SOFTWARE

DEVELOPMENT

263

about the different kinds of problems that may occur

from different perspectives. Furthermore, since the

VA controls the conversations it is not likely that

off-topic conversations will occur, which quite often

takes place when two or more students communicate

by chat. Students can therefore take better advantage

of their time than when working with other students.

In our future work we will test the framework

presented by collecting evidence that will allow its

effectiveness to be demonstrated.

We will guide our future efforts towards

answering the following research questions:

• Adequacy: Do the students understand the

purpose of the simulations? Do the students feel

that they have improved their skills in GSD?

• Time Requirements: How long do the students

need to complete the course and to finish the

deliverables? Do instructors appreciate a faster

organization and better performance in the GSD

courses?

• Usability: What problems occur during the

interaction with VAs when using VENTURE?

What is the students’ opinion of the feasibility of

the tool?

• Motivation: Do students feel motivated when

interacting with VAs? To what extent does a

student perceive the usefulness of the

framework?

With regard to the use of VAs, we intend to

study to what degree they induce a sense of social

presence in the students. This sense is increased by

the transmission of nonverbal cues showing

emotional states and gestures. In this respect, we

shall also study the following questions:

• Do students identify with their roles in the virtual

simulations?

• Do they consider that the experience of

interacting with VAs is realistic?

From the perspective of the instructors, we must

also study the feasibility of the framework for

designing and customizing the training materials and

scenarios. In this respect, the following questions

should be answered:

• What is the instructors’ opinion as regards the

use of the meeting designer? How long does it

take them to design a scenario?

• Evaluation and monitoring: What is their

perception of the monitoring and evaluation

facilities?

The students’ active participation in the

evaluation will be helpful in obtaining feedback with

which to improve the framework. We are also in the

process of preparing surveys, structured interviews

and in-situ observations.

Finally, we also wish to design and test practical

scenarios that will allow us to compare the

performance of students who have trained with our

framework with that of others of the same

characteristics who have not.

The results of the evaluations will eventually

help to design new training scenarios and to improve

the cultural and linguistic problems database.

ACKNOWLEDGEMENTS

This work has been funded by the PEGASO/MAGO

project (Ministerio de Ciencia e Innovación

MICINN and Fondos FEDER, TIN2009-13718-

C02-01). It is also supported by MEVALHE (HITO-

09-126) and ENGLOBAS (PII2I09-0147-8235),

funded by Consejería de Educación y Ciencia (Junta

de Comunidades de Castilla-La Mancha), and co-

funded by Fondos FEDER, as well as as well as

GlOBALIA (PEII11-0291-5274), Junta de

Comunidades de Castilla-La Mancha, Consejería de

Educación y Ciencia in Spain, and ORIGIN (IDI-

2010043 (1-5)) funded by CDTI and FEDER.

REFERENCES

Bradner, E. & Mark, G. 2002. Why distance matters:

effects on cooperation, persuasion and deception.

Proceedings of the 2002 ACM conference on

Computer supported cooperative work. New Orleans,

Louisiana, USA: ACM.

Casey, V. & Richardson, I. 2008. The Impact of Fear on

the Operation of Virtual Teams. Proceedings of the

2008 IEEE International Conference on Global

Software Engineering. IEEE Computer Society.

Cemile, S., Kathleen, S., Ferda, A., Robert, B., George, D.

& Lopez, V. 2009. Exploring the communication

behavior among global software development learners.

International Journal of Computer Applications in

Technology (Accepted).

Monasor, M. J., Piattini, M. & Vizcaíno, A. 2009.

Challenges and Improvements in Distributed Software

Development: A Systematic Review. Advances in

Software Engineering, 2009, 14.

Monasor, M. J., Vizcaíno, A., Piattini, M. & Caballero, I.

Year. Preparing students and engineers for Global

Software Development: A Systematic Review. In:

International Conference on Global Software

Development (ICGSE 2010), August 23-26, 2010 2010

Princeton, NJ, USA. 177-186.

ICSOFT 2011 - 6th International Conference on Software and Data Technologies

264